Purification and Characterization of a Mitochondrial Isozyme of C1-Tetrahydrofolate Synthase from Saccharomyces cerevisiae”

C1-Tetrahydrofolate synthase is a trifunctional poly- peptide found in eukaryotic organisms that catalyzes 10-formyltetrahydrofolate synthetase (EC 6.3.4.3), 5,lO-methenyltetrahydrofolate cyclohydrolase (EC 3.5.4.9), and 5,lO-methylenetetrahydrofolate dehy- drogenase (EC 1.5.1.5) activities. In Saccharomyces cereuisiae, C1-tetrahydrofolate synthase is encoded by the ADE3 locus, yet d e 3 mutants have low but detect-able levels of these enzyme activities. Synthetase, cy- clohydrolase, and dehydrogenase activities in an d e 3 deletion strain co-purify 4,000-fold to yield a single protein species as seen on sodium dodecyl sulfate-pol-yacrylamide gels. The native molecular weight of the isozyme (M. = 200,000 by gel exclusion chromatography) and the size of its subunits (Mr = 100,000 by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) are similar to those of Cl-tetrahydrofolate synthase. Cell fractionation experiments show that the isozyme, but not Cl-tetrahydrofolate synthase, is localized in the mitochondria. Genetic studies indicate that the isozyme is encoded in the nuclear genome. Peptide mapping experiments show that C1-tetra- hydrofolate synthase and the isozyme

ase activities are catalyzed by a trifunctional polypeptide, termed CI-THF synthase (4-9). CI-THF synthase from the yeast Saccharomyces cerevisiae is fairly representative of all these eukaryotic enzymes; it is a homodimer with a subunit molecular weight of 104,000 (5). In prokaryotes, these same activities are generally catalyzed by separate monofunctional enzymes which differ from Cl-THF synthase both in size and subunit composition (3).
The first evidence to indicate that there is an association between the synthetase, cyclohydrolase, and dehydrogenase activities in eukaryotes was genetic evidence in yeast. In S. cerevisiae, mutations that map to a single locus, the ADE3 locus, cause deficiencies in each of the three activities (10, 11). It was later shown that ADE3 is the structural gene for C1-THF synthase (12). It was originally reported that synthetase activity is "totally absent" in ade3 mutants (10). Yet, we (6) and others (11, 13, 14) have found low, but detectable, levels of synthetase, dehydrogenase, and cyclohydrolase activities in ade3 mutant strains. The presence of synthetase and dehydrogenase activities in yeast mitochondria has been reported (15). However, the nature of the enzyme or enzymes catalyzing these reactions was not examined. We report here the isolation and characterization of a mitochondrial isozyme of C1-THF synthase from S. cerevisiae. We call this isozyme "mitochondrial Cl-tetrahydrofolate synthase."

Materials
Common reagents were commercial products of the highest grade available. Culture media were purchased from Difco. Molecular weight markers for gel electrophoresis, hydroxylapatite (Bio-Gel HTP), and protein assay dye reagent were purchased from Bio-Rad. Percoll and Blue Sepharose were obtained from Pharmacia P-L Biochemicals. Protamine sulfate was purchased from Elanco Products Co. (Indianapolis, IN), Cm-cellulose from Whatman, Staphylococcus aureus V8 protease from Miles (Napervile, IL), a-chymotrypsin from Worthington, and Zymolyase from Kirin Brewing Co. (Tokyo).

Synthetase
New York) were designated p mutants.
Cells were grown aerobically at 30 "C. Growth was monitored by measuring turbidity at 600 nm with a Zeiss PMQ I1 spectrophotometer. Cells were grown in YPD medium (1% yeast extract, 2% bactopeptone, 2% glucose) to late log phase (ODm = 10) unless otherwise indicated. Extracts were prepared by disrupting washed cells with glass beads (0.45-mm diameter) in buffer containing 50 mM Tris-S04, pH 7.5, 10 mM KC1, 10 mM 2-mercaptoethanol, 1 mM PMSF. Homogenates were centrifuged 30 min at 16,000 X g. The supernatant fractions were used for enzyme and protein assays.

Gel Electrophoresis
The discontinuous buffer system described by Laemmli (30) was used for electrophoresis of polyacrylamide slab gels. Proteins and peptides were visualized by staining with Coomassie Brilliant Blue.

Isolation of Mitochondria
Cells were grown in medium containing 1% yeast extract, 1% bacto-peptone, and 1% galactose to late log phase. Cells were harvested, converted to spheroplasts with Zymolyase, and homogenized (31) in 2-3 volumes of ice-cold "breaking" buffer (0.6 M sorbitol, 10 mM Tris-C1, pH 7.4, 1 mM PMSF, 2 mM dithiothreitol, 8 mM ATP, 8 mM MgCl,). All subsequent steps were carried out at 4 "C. The homogenate was centrifuged at 1400 X g for 5 min, and the supernatant fraction was then centrifuged at 9600 X g for 10 min. The resulting pellet was resuspended in breaking buffer and centrifuged as above. The crude mitochondria (the washed pellet) were further purified on a 25% Percoll gradient formed in situ in 0.25 M sucrose, 10 mM Tris-C1, pH 7.4, 1 mM PMSF, 2 mM dithiothreitol, 8 mM ATP, and 8 mM MgC1,. Centrifugation was carried out at 64,000 X g for 30 min. An orange-brown band which sedimented toward the bottom of the gradient was collected, and the Percoll was removed from the suspension by high speed centrifugation (106,000 X g for 90 rnin). The mitochondria were separated from the hard pellet formed by the Percoll by rinsing with breaking buffer and were recovered by centrifuging at 9,600 X g for 10 min. Mitochondria were stored at -80 "C prior to extraction. Extracts were prepared as described above.

Peptide Mapping Experiments
One-dimensional peptide mapping was performed as described (32). Purified enzymes (10 pglwell) were mixed with either S. aureus V8 protease or a-chymotrypsin. The mixtures were incubated at room temperature for 30 min, and the products were separated on a 12.5% SDS-polyacrylamide gel.

Antibody Preparation and Immunotitration Experiments
Antisera against yeast CI-THF synthase and mitochondrial CI-THF synthase were raised in rabbits (33), and IgG was purified by ammonium sulfate precipitation and DEAE-cellulose chromatography (34). Purified enzymes (0.91 pg/ml in 50 mM K-MES, pH 6.5,10 mM KC1, 3% bovine serum albumin) and various dilutions of antiserum were incubated at 37 "C for 30 min and then stored on ice for 2 h. Immune complexes were precipitated from the solution as previously described (35), and aliquots of the supernatant fraction were assayed for IO-formyl-THF synthetase activity.

Amino Acid Sequence Analysis
The amino terminus of mitochondrial C1-THF synthase (residues 1-40) was sequenced by Dr. Stanley C. Rall, Jr., Gladstone Foundation Laboratories, University of California, San Francisco. The purified protein was applied to a Beckman 890M Sequencer and the phenylthiohydantoin-derivatized amino acids were identified by high pressure liquid chromatography on a Beckman model 332 HPLC equipped with an Ultrasphere ODS reverse phase column.

Purification of Mitochondrial CI-THF Synthase
Cell Growth and Storage-Cells of yeast strain 3-5281 were grown in YPD medium supplemented with adenine (20 mg/l), harvested at late log phase, and stored at -80 "C for up to 6 months without loss of activity.
Cell Extract-Cells (1 kg) were thawed and suspended to 50 g/100 ml in standard enzyme buffer (50 mM Tris-SO+, pH 7.5,lO mM KCl, 10 mM 2-mercaptoethanol) containing 1 mM PMSF. The suspension was homogenized by five passes through a Manton-Gaulin homogenizer at a pressure of 8,000 p.s.i. The temperature of the suspension was maintained below 25 "C during homogenization. All subsequent steps were performed at 0-4 "C. The homogenate was centrifuged at 16,000 X g for 30 min and the supernatant solution was decanted through a pad of glass wool.
Protamine Sulfate Fractionation-A solution of protamine sulfate, 1% in standard enzyme buffer, was added to the extract to a final concentration of 0.2%. After stirring for 30 min, the suspension was centrifuged at 16,000 X g for 10 min, and the supernatant solution was collected. PMSF was added to 1 mM final concentration.
Ammonium Sulfate Fractionation-Solid ammonium sulfate (29.1 g/100 ml) was added to the protamine sulfate supernatant to achieve 50% saturation (36). The suspension was stirred for 30 min and centrifuged at 16,000 X g for 10 min. The supernatant solution was collected, adjusted to 65% saturation by adding more solid ammonium sulfate (9.2 g/lOO ml), stirred for 30 min, and centrifuged as before. The pellet was resuspended in a minimal volume of K-MES buffer (20 mM K-MES, pH 6.5, 10 mM 2-mercaptoethanol) containing 10 mM KC1. PMSF was added to 1 mM final concentration here and to the subsequent pooled column fractions.
Cm-cellulose Chromatography-The ammonium sulfate fraction was dialyzed extensively against K-MES buffer containing 10 mM KC1, centrifuged at 28,000 X g for 10 min to remove particulate material, and applied to a Cm-cellulose column (5.0 X 20 cm) equilibrated in K-MES buffer containing 10 rnM KCl. After washing with 5-column volumes of K-MES buffer containing 0.04 M KC1, the column was developed with a linear gradient (5-column volumes) between 0.04 and 0.4 M KC1 each in K-MES buffer. 10-Formyl-THF synthetase activity eluted in a single peak at a conductivity of 9 mmho, and fractions containing activity were pooled.
Heparin-Agarose Chromatography-The pooled fraction from the Cm-cellulose column was applied to a heparin-agarose column (1.5 X 10 cm) equilibrated with K-MES buffer containing 0.15 M KCI. After washing with 10-column volumes of the equilibration buffer containing 15 mM ATP, the column was developed with a linear gradient (10-column volumes) between 0.15 and 0.8 M KC1 each in K-MES buffer. 10-Formyl-THF synthetase activity eluted in a single peak at a conductivity of 18 mmho, and fractions containing activity were pooled.
Blue Sephrose Chromatography-The pooled fraction from the heparin-agarose column was applied to a Blue Sepharose column (1.5 X 7.0 cm) equilibrated with K-MES buffer containing 0.5 M KCl. After washing with 10-column volumes of the equilibration buffer, the column was developed with a linear gradient (10-column volumes) between 0.5 and 2.5 M KC1 each in K-MES buffer containing 10% glycerol. 10-Formyl-THF synthetase activity eluted in a broad peak, and fractions containing activity were pooled.
Hydroxylapatite Chromatography-The pooled fraction from the Blue Sepharose column was dialyzed against 0.1 M sodium phosphate buffer, pH 6.5,0.05 M KCl, 10 mM 2-mercaptoethanol and applied to a hydroxylapatite column (1.0 X 3.5 cm) equilibrated with the dialysis buffer. After washing with 10-column volumes of the dialysis buffer, the column was developed with a linear gradient (20-column volumes) between 0.1 and 0.6 M sodium phosphate buffer, pH 6.5, each in 0.05 M KCl, 10 mM 2-mercaptoethanol. 10-Formyl-THF synthetase activity eluted in a single peak at a conductivity of 12 mmho, and fractions containing activity were pooled. The purified enzyme was concentrated, dialyzed against K-MES buffer containing 25 mM KCI, 50% glycerol, 1 mM PMSF, and stored at -20 "C.

RESULTS
Purification of Mitochondrial C1-THF Synthase-In S. cereuisiae, Cl-THF synthase is encoded by the ADE3 locus. However, extracts from an ade3 mutant strain (3-5281) had approximately 10% of the synthetase activity, 30% of the dehydrogenase activity, and 6% of the cyclohydrolase activity as an ADE3 wild-type strain (M16-14C) ( Table I). This particular mutant strain carries an extensive deletion at the ADE3 locus and lacks Cl-THF synthase or any kind of derivative of Cl-THF synthase (33,37). The "residual" activities that were found in the ade3 mutant could be accounted for in two ways. First, there could exist an isozyme of C1-THF synthase catalyzing all three of the activities. Alternatively, the activities could be catalyzed by three separate monofunctional enzymes. To distinguish between these pos- sibilities, we chose to purify the synthetase activity from the ade3 deletion strain and examine the catalytic activities of the purified protein. Extracts were treated with protamine sulfate and ammonium sulfate and then applied to Cm-cellulose, heparin-agarose, Blue Sepharose, and hydroxylapatite. We achieved a 4000-fold purification of the synthetase activity with a 16% yield (Table 11). SDS-polyacrylamide gels of the purified sample reveal a single band when stained with Coomassie Brilliant Blue (Fig. 2). We performed densitometry at 560 nm of overloaded gels (10 pg protein/lane). The results demonstrate that the preparation is at least 91% pure (data not shown). The dehydrogenase and cyclohydrolase activities that were present in the crude extract co-purified with the synthetase activity (Table 11). These data suggest that the purified protein is trifunctional and catalyzes the same reactions as Cl-THF synthase. The purified protein is not encoded by the ADE3 gene, thus we can refer to this protein as an isozyme of CI-THF synthase. SDS-polyacrylamide gel electrophoresis showed that the purified isozyme has a molecular weight of 100,000 (Fig. 2). Gel exclusion chromatography on Sephacryl $3-300 showed that the native molecular weight of the isozyme is approximately 200,000 (data not shown). These data suggest that the isozyme is composed of two identical subunits, as is C1-THF synthase.
Subcellular Localization of Mitochondrial C1-THF Synthase-The isozyme is present in ade3 mutants, but it cannot satisfy the requirements for adenine and histidine that are characteristic of most d e 3 mutants (38). This suggests that Cl-THF synthase and the isozyme may have different cellular functions. Zelikson and Luzzati (15) suggested that the synthetase and dehydrogenase activities in ade3 mutants are found in the mitochondria. However, these investigators did not examine the purity of their mitochondria, and we could not distinguish between a mitochondrial localization and a vacuolar localization for these activities using their procedure for mitochondrial isolation. We isolated mitochondria from strains M16-14C and 3-5281 using a modified procedure and measured synthetase activity in the mitochondrial extracts. To test the integrity and the purity of our isolated mitochondria, we also measured the activities of a number of enzymes that mark various subcellular compartments. In both strains, the mitochondrial marker enzymes (fumarase and cytochrome b,) were enriched in isolated mitochondria, whereas the enzymes that mark the vacuoles, the cytoplasm, the endoplsmic reticulum, and the secretory vesicles were not (Fig. 3). Synthetase was enriched in the mitochondria isolated from the    d e 3 deletion strain but not in those isolated from the ADE3 wild-type strain (Fig. 3). These results indicate that the isozyme is localized in the mitochondria, whereas C,-THF synthase is not.

Localization of Gene Encoding Mitochondrial C,-THF Syn-
t h e -T h e mitochondrion has its own genome that encodes rRNAs, tRNAs, plus a small number of mitochondrial proteins (39). Having shown that the isozyme is localized in the mitochondria, we wanted to determine whether it is encoded in the nuclear or the mitochondrial genome. The mitochondrial genome of a yeast p mutant is either absent or extensively deleted (40), thus a p mutant cannot synthesize mitochondrially encoded proteins, We isolated 10 p mutants from strain 3-5281 and found synthetase and dehydrogenase activities in extracts from each of them (Table 111). Although there was significant variability in the levels of activity among the mutant strains, the ratio of the specific activity of synthetase to that of dehydrogenase remained relatively constant and was comparable to that of the parent strain. These data indicate that mitochondrial Cl-THF synthase is encoded in the nucleus.
Peptide Maps of C1-THF Synthase and Mitochondrial C1-THF Synthase-Mitochondrial C1-THF synthase, being a nuclearly encoded mitochondrial enzyme, is presumably initially synthesized with an amino-terminal extension that directs the protein to the organelle (41). Thus, we would expect to find differences between mitochondrial Cl-THF synthase and C1-THF synthase at the level of their DNA sequences. However, the two mature proteins could be structurally identical. One procedure we used to examine the structural relationship between the isozymes was peptide mapping. This method is considered to be a stringent test for protein identity. It will also reveal close structural relationships (e.g. precursorproduct) among proteins. Purified mitochondrial C1-THF synthase and Cl-THF synthase were partially digested with proteases, and the peptide products were separated on SDSpolyacrylamide gels. The peptide banding patterns showed no striking similarities (Fig. 4), which indicates that C1-THF synthase and mitochondrial Cl-THF synthase are not structurally identical. The dissimilarities in the banding patterns probably reflect differences between the two proteins at the level of their primary sequences.
Immunotitration of Cl-THF Synthase and Mitochondrial Cl-THF Synthase-A second criteria we used to assess the structural relatedness of these isozymes was to determine whether they would cross-react immunologically. To quantitate cross-reactivity, we did immunotitration experiments by incubating the purified enzymes with polyclonal antisera directed against the purified proteins, precipitating the immune complexes and assaying the synthetase activity which remained in the supernatant fractions. Antiserum against mitochondrial C,-THF synthase reacted with both mitochondrial CI-THF synthase and Cl-THF synthase; however, the immunotitration curves were distinct (Fig. 5A). The reciprocal experiment with antiserum against C1-THF synthase showed similar results (Fig. 5B). The pre-immune sera did not affect synthetase activity in these experiments (data not shown). These results are consistent with the conclusion that the two isozymes are not structurally identical, but they also suggest that there are conserved features between the two proteins probably at the level of their secondary or tertiary structures.
Amino Acid Sequence Analysis of Mitochondrial Cl-THF Synthase-To examine the structural relatedness of the isozymes using a more stringent criteria, we compared a partial amino acid sequence of mitochondrial C,-THF synthase (residues 1-40) with the amino acid sequence deduced from the nucleotide sequence of the gene encoding C,-THF synthase (37). The sequence corresponding to the amino terminus of Cl-THF synthase aligned with the amino-terminal sequence of mitochondrial Cl-THF synthase (Fig. 6). In this region, there is 40% identity between the two sequences. Cl-THF synthase has two functionally independent domains (5, 42). Synthetase activity is catalyzed on the carboxyl-terminal domain, and dehydrogenase and cyclohydrolase activities are catalyzed on the amino-terminal domain. Although we have not shown that mitochondrial CI-THF synthase also has a two-domain structure, the alignment of the two sequences at their amino termini suggests that the arrangement of the activities within the protein is the same for both isozymes.

DISCUSSION
We have purified a mitochondrial isozyme of CI-THF synthase from S. cereuisiae which we call mitochondrial Cl-THF synthase. Mitochondrial C1-THF synthase and Cl-THF synthase are encoded by separate nuclear genes. The isozymes are not identical, but they have similar physical properties, and they are structurally related. Mitochondrial CI-THF synthase is present in ade3 mutants that lack CI-THF synthase. The observation that the isozyme cannot supply one-carbon units for the synthesis of purines in these mutants suggests that mitochondrial Cl-THF synthase and Cl-THF synthase may have different cellular roles.
There are several examples of isozymes that are differentially distributed between the mitochondria and the cytoplasm. These include a number a tricarboxylic acid cycle enzymes and enzymes involved in amino acid transamination, NADPH generation, gluconeogenesis, and energy transfer (43). In most of these cases, it has been shown that the two forms of these enzymes perform different metabolic functions. For example, the mitochondrial form of malate dehydrogenase oxidizes malate in the citric acid cycle, whereas the cytoplasmic form reduces oxaloacetate to supply malate for lipogenesis and malate-aspartate shuttle reactions. The different roles of the isozymes are usually manifested by differences in their catalytic properties. The kinetic parameters of mitochondrial C1-THF synthase have yet to be determined and compared to those of CI-THF synthase (4).
Another explanation for the different roles of the isozymes is that the products of the enzymes are required in two cellular compartments but are not transported across the membrane boundaries. In fact, there is evidence that folate coenzymes are not transported across the mitochondrial membrane (44, 45), although folate derivatives (46-49) and folate-dependent enzymes (50-54) are found in mitochondria. This raises the question of the source of the folates and the one-carbon groups required for one-carbon metabolism in the mitochondria. It has been proposed that one-carbon units are shuttled between the cytoplasm and the mitochondria by cytosolic and mitochondrial isozymes of serine hydroxymethyltransferase (44). This reactions requires THF and generates 5,lO-methylene-THF. In this scheme, C1-THF synthase and mitochondrial Cl-THF synthase would convert the 5,lO-methylene-THF generated in these reactions to other THF intermediates required for various biosynthetic pathways in the cytoplasm and the mitochondria.
Reduced folates are required for mitochondrial function (55), possibly because mitochondrial protein synthesis is dependent on the formylation of f-Met-tRNAmet (56). Mitochondrial CI-THF synthase may supply the substrate, 10formyltetrahydrofolate, for this reaction. We are now constructing mutations in the gene encoding mitochondrial C1-THF synthase to assess the physiological role of this enzyme.